Without restoration, modernized bogs will likely have a delayed successional progression. Noticeable differences in biomass occurred between the sites, with restored sites having significantly higher herbaceous and woody biomass. Although species richness in 1m2 quadrats was not markedly increased by restoration, H' indices for species diversity increased, perhaps by reducing dominance of the species suited to the uniformity of the former agricultural habitat. The heterogeneity created by the restoration activities has allowed for biomass to accumulate more rapidly than in the control or the peat-based bog, and although not significantly different, there is a measurably thicker hydric layer in the soil of the restored sites. These changes in vegetative communities and soil formation are essential for the ecosystem to regain functions of wildlife and plant habitat value, restoring the water holding capacity of the system to resist drought and assist in flood and erosion control. Compared with upland farms characterized by a homogenous landscape, the diverse habitats within cranberry farms create heterogeneity with various levels of anthropogenic disturbance (Wen 2010). This characteristic is particularly true for agricultural wetlands of the FPP. Our finding of lower species richness in the peat-based bog compared to modernized bogs is likely due to the fact that while cranberries were being grown, some native plant species coexisted within the intact wetland. After agriculture ceased and cranberry desiccated in winter, co-existing native wetland species gained dominance. In modernized bogs, however, cranberries were maintained in monoculture. After agriculture ceased, cranberry plants died and there were higher levels of species introduction and colonization. We found that the FPP plant community is closely linked to the degree of agricultural manipulation or restoration of the vegetation and soil. Through reintroducing a dynamic hydrology, mounded topography, and hydric soil, restoration activities instituted by the NJCF are accelerating succession of the modernized bogs into swamps. In addition to the restoration activities reported here, the NJCF is actively reintroducing Atlantic white-cedar (Chamaecyparis thyoides) and native forbs to many of the modernized cranberry bogs after the microtopography is restored, although planting activity did not occur yet in the sites selected for this study. The planting of red maple is not necessary, since that species is being naturally recruited on the site by neighboring intact swamps, as demonstrated by this research. The development of vegetative communities resulting from the NJCF activities is essential in order for the ecosystem to regain functions of wildlife and plant habitat value and restoring the water holding capacity of the system.
References Ahn, C. and S. Dee. 2011. Early development of plant community in a created mitigation wetland as affected by introduced hydrologic design elements. Ecological Engineering 37: 1324–1333. Eck, P.1990. The American Cranberry. New Brunswick, NJ: Rutgers University Press. Forman, R.T.T 1998. Pine Barrens: Ecosystem and Landscape. New Brunswick, NJ: Rutgers University Press. Good, R. E. and N.F. Good. 1984. The Pinelands National Reserve: An Ecosystem Approach to Management. BioScience 34(3): 169–173 Mylecraine, K.A., G.L. Zimmermann, R.R Williams, and J.E. Kuser. 2004. Atlantic white-cedar wetland restoration on a former agricultural site in the New Jersey Pinelands. Ecological Restoration 22(2): 92–98. Peet, R.K., T.R. Wentworth, and P.S. White. 1998. A flexible, multipurpose method for recording vegetation composition and structure. Castanea 63(3): 262–274. Procopio, N.A. 2010. Hydrologic and morphologic of variability in streams with different cranberry agriculture histories. Journal of the American Water Resources Association 46: 527–540. Reiser, J. 2014. New Jersey State Atlas: 1930s Aerial Photography. Trenton, NJ: New Jersey Department of Environmental Protection and Office of Information Technology. (http:// njstateatlas.com/1930/Geo: 39.772444, -74.529383 USNG: 18S WK 4030 0260). United States Department of Agriculture, Natural Resources Conservation Service. 2010. Field Indicators of Hydric Soils in the United States, Version 7.0. L.M. Vasilas, G.W. Hurt, and C.V. Noble (eds). Washington, DC: USDA, NRCS, in cooperation with the National Technical Committee for Hydric Soils. Wen, A. 2010. Ecological functions and consequences of cranberry (Vaccinium macrocarpon) agriculture in the pinelands of New Jersey. PhD dissertation, Rutgers University.
Soil Amendment Increases Tree Seedling Growth but Reduces Seedling Survival at a Retired Gravel Mine
Nate Hough-Snee (Utah State University Ecology Center and Department of Watershed Sciences, Logan, UT), and Rodney Pond (corresponding author: University of Washington School of Environmental and Forest Sciences and Center for Urban Horticulture Box 354115, Seattle, WA 98195-2100, rodney.
[email protected]).
R
estoring forest vegetation within denuded settings requires identifying the abiotic factors that limit plant establishment (Bradshaw 1997, Whisenant 1999). In heavily disturbed sites such as quarries and gravel pits that have been denuded of native soils, degraded soil processes may limit planted tree seedling survival or growth
September 2014 ECOLOGICAL RESTORATION 32:3 • 231
Table 1. Mean soil carbon and C to N ratio differed between amended and unamended plots at both 0–15 cm and 15–30 cm depths. Soil moisture and nitrogen were higher at shallow depths in the amended plots, while bulk density did not differ between amended and unamended plots. Letters in 2002 and 2005 data indicate significant differences (ANOVA; F1,36 > 4.12, p < 0.05 for all tests) between amendment treatments within a given soil depth and year. Due to limited sampling, no statistics were performed on pre-restoration soil monitoring. Soil Parameter Carbon Nitrogen C to N Ratio Bulk Density % Soil Moisture
Depth
Gravel Mine
0–15cm 15–30cm 0–15cm 15–30cm 0–15cm 15–30cm 0–15cm 15–30cm 0–15cm 15–30cm
0.15 — 0 — 48 — — — — —
Pre-Restoration Volunteer Mature Reference Reference 2.37 0.62 0.08 0.02 30 29 — — — —
2.20 1.18 0.08 0.04 28 24 — — — —
(Williamson et al. 2011). To improve vegetation establishment on denuded sites, soil amendments are commonly used to manipulate soil fertility (Biederman and Whisenant 2011a, Hough-Snee et al. 2011a), introduce organic matter , soil microorganisms, or propagules (Sinnett et al. 2008, Hough-Snee et al. 2012), or to create heterogeneous microhabitats (Biederman and Whisenant 2011b, Hough-Snee et al. 2011b) that improve plant survival. In forest restoration, amendments that facilitate soil development can lead to increased plant survival and growth (Ortiz et al. 2011). Specifically, amendments that increase soil carbon and nitrogen have been shown to improve planted tree seedling growth (Wilson-Kokes et al. 2013). Soil amendments for denuded sites are typically designed to increase planted tree growth so that seedlings may outcompete early seral vegetation and survive to maturity (Bradshaw 1997). In this study we examined how soil amendments change soil properties at a highly disturbed site and how these amendments impact the growth and survival of three early successional tree species. We tested two sets of hypotheses: 1. Soil amendment will increase soil carbon and nitrogen, C to N ratio, soil moisture, and decrease soil bulk density. 2. The amendment-driven increase in soil fertility will increase the survival and growth of planted black cottonwood (Populus balsamifera), red alder (Alnus rubra) and Douglas fir (Pseudotsuga menziesii) seedlings relative to unamended seedlings.
The restoration site was a 1.7-ha retired gravel mine located on an alluvial terrace near Goodell Creek, a tributary to the Skagit River (Washington State, USA, elevation: 162 m). Gravel operations ceased in 1990 when the mine and surrounding area were incorporated into North Cascades National Park. The surrounding matrix consists of mature Douglas fir- and western hemlock- (Tsuga
2002
2005
Amended
Unamended
Amended
Unamended
40.41a 11.86a 0.49a 0.39ns 87.68a 30.86a 0.65ns 1.66ns 17.39a —
4.3b 5.22b 0.26b 0.36ns 16.27b 13.79b 1.28ns 2.41ns 2.36b —
25.82a 18.33a 0.81a 0.51a 37.82a 40.42a 0.67ns 1.71ns 19.17a 8.72 ns
3.21b 5.67b 0.22b 0.76b 7.09b 11.12b 1.17ns 2.25ns 6.17b 6.17ns
heterophylla) dominated conifer forest in uplands and black cottonwood-, red alder-, and western red cedar- (Thuja plicata) dominated riparian forest. Soil parameters were sampled within both forest types and the gravel mine to identify soil conditions prior to restoration (Table 1). The primary restoration objective was to use amendments to establish soil properties that facilitate early-successional, coniferous-deciduous forest stand development. Temperate forests of Washington’s western Cascades are generally low in available nitrogen, so amendments were designed to raise soil organic matter content and moisture retention capacity without increasing N mineralization that would favor competitive ruderal weed establishment. Prior to amendment application, the entire site was graded to a ~7% grade and stockpiled sandy loam aggregate was evenly spread to a depth of 15 cm. Soil amendment consisted of a secondarily digested paper pulp sludge stabilized with fly ash (Smukler 2003). The amendment had high initial nitrogen content, so the carbon to nitrogen (C to N) ratio was increased prior to application by adding partially decomposed alder sawdust (36% C, 0.29% N, C to N ratio = 125; Smukler 2003). In the summer of 2001, the amendment was spread across two 0.45 ha blocks and tilled into the sandy loam topsoil (15 cm deep) over the mine’s semi-compacted subsoil. Two 0.45-ha blocks remained unamended and were not tilled. Thirty-six 20 m2 circular experimental plots were created, nine per block. We measured soil percent carbon, percent nitrogen, moisture, and bulk density, and calculated the C to N ratio at plot centers in early summer of 2002 and 2005 (methods discussed in Smukler 2003). Soils were sampled at 0–15 cm and 15–30 cm depths to identify amendment effects across depths. As part of an additional experiment on promoting revegetation through seed rain recruitment (Pond 2005), three mulch treatments were randomly and evenly applied to plots within amended and unamended blocks,
232 • September 2014 ECOLOGICAL RESTORATION 32:3
Figure 1. Frequency plots of mortality by year, species, and soil amendment show that amended individuals were more likely to die than unamended individuals. All species had similar mortality rates, but there were twice as many Douglas fir (DF) seedlings as red alder (RA) and black cottonwood (BC).
woodchips (chipped red alder bark), straw (weed-free), and a control. Mulches were applied to plot surfaces (2–3 cm deep) within each block to influence seed capture, retention, and germination, but not rooting zone soil properties. Bare root seedlings (1–0) of red alder, black cottonwood, and Douglas fir were planted in November 2001. Four Douglas fir, two cottonwood, and two alder were planted within each circular plot at evenly spaced, random compass headings. Tree height was measured at planting and ranged from 40 to 60 cm for all species. Height was measured from the soil surface to the end of the main stem terminal bud.
Tree survival and height were surveyed in 2002, 2004, and 2005. Growth was calculated as the difference in heights between each measurement year and 2001 height at planting. If tissue dieback occurred then measured heights less than those at the time of planting were recorded as negative measurements. Initially we tested for the effects of amendment and mulch treatments on soil properties using two-way ANOVA. There were no significant mulch effects on soil properties at either year or depth, so this effect was removed and one-way ANOVA was performed on soil properties
September 2014 ECOLOGICAL RESTORATION 32:3 • 233
2002
2004
2005
400
−−
Growth (cm)
300
Amendment Unamended
200
−− • •
100 •
0
−− •
−− ••
RA RA
• •• •
•• −− −− −− •• −− • •
−−
−− −− −−
•
−− −− −− −− •
• •
−− −−
•
•
BC BC DF DF
Amended
RA RA
BC BC DF DF
Tree species
RA RA
BC BC DF DF
Figure 2. Box and whisker plots of each species’ growth between 2001 and 2002, 2004, and 2005. Amended red alder (RA) grew the most of all species and treatments across all growth intervals. Amended black cottonwood (BC) and Douglas fir (DF) outgrew their unamended counterparts. Black bars (—) indicate the mean change in height for a given species within a given treatment and year.
for amendment. We used binomial regression to identify differences in tree survival between soil amendment treatments and ANOVA to test for the effects of amendment on surviving seedling growth. Each species’ growth was analyzed independently between 2001–02, 2001–04, and 2001–2005. Because we were interested in the effect of amendment on soils and tree survival and growth, when mulch effects were not significant we removed this term from models, testing only for amendment effects. Amendment elevated soil C, N, and C: N ratio above the levels found in volunteer and mature reference forests and the pre-restoration gravel mine. Soil carbon and C:N ratio were higher in amended plots than unamended plots at both shallow and deep depths in both 2002 and 2005 (Table 1). Nitrogen was higher in amended plots at shallow depths in both years. Bulk density was lower in amended plots than in unamended plots although not statistically significant in both years. Soil moisture was higher in amended plots than in unamended plots in both years. There were no significant mulch effects on soil properties in either year or depth. Soil treatment means and statistical results are presented in Table 1. More unamended cottonwood seedlings died than amended seedlings in 2002. In 2004 and 2005 more unamended cottonwood seedlings survived than amended seedlings (Binomial regression; 2004: Z = 2.76, p = 0.006; 2005: Z = 3.23, p = 0.001; Figure 1). Amended Douglas fir seedlings experienced higher mortality than unamended seedlings over the study duration (Binomial regression;
2002: Z = 2.04, p = 0.042; 2004 and 2005: Z = 2.56, p = 0.011; Figure 1). Red alder mortality did not differ between amended and unamended seedlings in 2002, but was higher in amended seedlings in 2004 and 2005 (Binomial regression; 2002: Z = 1.63, p = 0.104; 2004: Z = 1.70, p = 0.089; 2005: Z = 2.00, p = 0.046; Figure 1). Mulch treatment was not significant in any survival models. Cottonwood seedling growth was higher in amended plots across the 2001–02 interval (ANOVA; F = 7.39, p = 0.009). Amendment did not significantly improve cottonwood growth between the 2001–04 and 2001–05 growth intervals (Figure 2). Douglas fir growth was higher in amended plots than in unamended plots between 2001–04 (ANOVA; F = 6.51, p = 0.013) and 2001–05 (F = 5.25, p = 0.025), but did not differ between 2001–02. Red alder growth was significantly higher in amended plots across all intervals (ANOVA; 2001–02: F = 32.08; p < 0.001; 2001–04: F = 28.84, p < 0.001; 2001–2005: F = 53.67, p < 0.001; Figure 2). Mulch was also a significant term in the model for alder growth between 2001–02 (ANOVA; F = 3.484, p < 0.037). Although straw mulch increased alder growth within amended plots in 2002, no other mulch treatment affected growth or survival, therefore we do not present mulch-amendment pairwise comparisons. Amendment increased soil nutrition and moisture, resulting in increased growth of surviving tree seedlings. For all three species, amended plots had higher seedling growth than unamended plots. While these results confirmed our initial hypotheses, survival results were
234 • September 2014 ECOLOGICAL RESTORATION 32:3
less intuitive. Soil amendment actually increased plant mortality over the four-year study duration. This could be due to excessively low soil bulk density coupled with increased nutrition that promoted rapid aboveground growth but only shallow rooting. The average bulk density of unamended soils was twice that of amended soils in both years across the soil column. Low amendment bulk density likely resulted from the use of sawdust to balance carbon and nitrogen ratios. At low bulk densities, macropores (airspaces that don’t allow capillary water movement in the soil column) increase, reducing the soil’s capacity to retain water near the soil surface during dry conditions. Although soil amendments retained moisture at shallow depths under wet conditions, it is possible that the loose texture and low density encouraged rapid drying in young seedlings’ rooting zones during summer drought. Soil drying would be most disadvantageous to cottonwood and alder that have high transpiration and generally need consistent moisture to survive. Another explanation for higher mortality in amended plots is that soil amendment increased growth and competition among planted seedlings. Competitive exclusion may have most adversely affected relatively slow-growing, shade intolerant Douglas fir, especially in amended plots in 2004 and 2005. Red alder, a nitrogen-fixing tree, grew rapidly and 100 times taller than Douglas fir in amended plots. This suggests that species’ life history strategies will influence how planted vegetation communities respond to soil enrichment in stressful environments. When the primary filter that shapes community assembly is physical (i.e. soils), then using amendments to improve physical properties will allow vegetation to establish. However, once vegetation has established, biotic filters (i.e. competition) will shape planted vegetation survival and growth. At our site, faster-growing alder and cottonwood may have shaded Douglas fir, increasing mortality and slowing growth. When using amendments that give species with competitive strategies a growth advantage, successional management may be necessary to increase slow-growing species survival. Thinning adjacent woody vegetation, weeding, or using plastic mulches to prevent competing vegetation from establishing are all viable options to ensure that planted seedlings survive and grow to shape future community assembly.
References Biederman, L.A. and S.G. Whisenant. 2011a. Amendment placement directs soil carbon and nitrogen cycling in severely disturbed soils. Restoration Ecology 19:360–370. Biederman, L.A. and S.G. Whisenant. 2011b. Using mounds to create microtopography alters plant community development early in restoration. Restoration Ecology 19:53–61. Bradshaw, A. 1997. Restoration of mined lands using natural processes. Ecological Engineering 8:255–269. Hough-Snee, N., J.D. Bakker and K. Ewing. 2011a. Long-term effects of initial site treatment on fescue in a novel prairie ecosystem (Washington). Ecological Restoration 29:14–17. Hough-Snee, N., A.L. Long, L. Jeroue and K. Ewing. 2011b. Mounding alters environmental filters that drive plant community development in a novel grassland. Ecological Engineering 37:1932–1936. Hough-Snee, N., Long, A.L. and R.L. Pond, 2012. Passive soil manipulation influences the successional trajectories of forest communities at a denuded former campsite. Ecological Restoration 30:9–12. Ortiz, O., G. Ojeda, J.M. Espelta and J.M. Alcañiz. 2011. Improving substrate fertility to enhance growth and reproductive ability of a Pinus halepensis afforestation in a restored limestone quarry. New Forests 43:365–381. Pond, R.L., 2005. Low elevation riparian forest restoration on a former gravel mine, North Cascades National Park USA: Native plant germination, growth and survival in response to soil amendment and mulches. Masters Thesis. University of Washington, Seattle, WA. Sinnett, D., J. Poole and T.R. Hutchings. 2008. A comparison of cultivation techniques for successful tree establishment on compacted soil. Forestry 81:663–679. Smukler, S.K., 2003. Optimizing the use of residuals as soil amendments for ecological restoration. Masters Thesis. University of Washington, Seattle, WA. Whisenant, S.G. 1999. Repairing Damaged Wildlands a Process-oriented,Landscape-scale Approach. Cambridge, UK: Cambridge University Press. Williamson, J.C., E.C. Rowe, P.W. Hill, M.A. Nason, D.L. Jones and J.R. Healey. 2011. Alleviation of both water and nutrient limitations is necessary to accelerate ecological restoration of waste rock tips. Restoration Ecology 19:194–204. Wilson-Kokes, L., C. DeLong, C. Thomas, P. Emerson, K. O’Dell and J. Skousen. 2013. Hardwood tree growth on amended mine soils in West Virginia. Journal of Environment Quality 42:1363–1371.
Acknowledgements We thank North Cascades National Park and the National Park Service Challenge Cost Share Program for funding, Gina Rochefort, Warren Gold, and Kern Ewing for guidance throughout the project and Sean Smukler for soil sampling design and collection.
September 2014 ECOLOGICAL RESTORATION 32:3 • 235
